« first day (309 days earlier)      last day (2185 days later) » 

00:01
@Dennis I think QuadR and QuadS (without spaces) are even catchier. But you should probably have a parens saying (Dyalog APL) in the display names.
@Dennis Even though the code uses , Classic's ⎕FIX is clever enough to translate that to ` ⎕U2360 ` when loading, so it can be represented in the SBCS.
@Adám I'm not sure about the parens part. I also don't write Python (gcc) just because it was compiled with gcc.
@Dennis OK, then leave it out. I hope I won't give you too many problems with the Dyalog dependency.
Alright, as soon as I get to my computer, I'll try Hello World again.
@Dennis I guess I should remove the TIO listings from the README.md, no?
Probably. What you could do (if you want to) is include the wrappers in the repo, in executable format and taking the Dyloag directory as i put. That way, if the wrapper ever has to change, you can do it in the repo.
00:16
@Dennis How do I see what you put?
@Dennis Also, now I think about it, their links should probably go to github, and the README should have links to Dyalog's documentation.
Yes, that would be better. I've linked to GitHub on my server.
@Dennis Great. I'll update my file.
@Adám Give me a minute. I'll make a draft.
00:30
#!/usr/bin/env bash

export DYALOG=$1
export MAXWS=128M WSPATH=$DYALOG/ws

cat <<- . | "$DYALOG/dyalog" -script
	⎕PW←9999
	{}2⎕FIX'file://$(dirname "$BASH_SOURCE")/Run.dyalog'
	'$2'('R'Run '${@:4}')'$3'
.
Something like this. I could invoke it as quadr /opt/mdyalog/15.0/64/unicode .code.tio .input.tio "$@", and so could everyone else with different filenames.
What does <<- do?
@Dennis neat. That's make it easier to upgrade APL (soon!) too. I'll add the two files to my repo as quadr and quads
@Phoenix << is a heredoc. The - makes it ignore leading tabs.
Ah, I knew what << was but wasn't sure about the -.
 
9 hours later…
09:56
How does one use a language where the source is raw bytes on TIO?
10:48
Now I'm wondering if I should add a language to TIO before I have tested it one time.
In a vacuum I would have guessed that is not a good thing to do, but it seems to be the standard practice.
11:00
@Dennis I've noticed that some languages like cjam, pyth etc. are encoded as sbcs while they're really encoded as utf8...
@EriktheOutgolfer They can use both, but TIO, running in a browser, must use UTF-8 to enable human reading and writing.
umm...I thought cjam uses utf8 as its encoding...and tio mistakens this as sbcs
for example, "ḤḢý" wouldn't be 5 bytes in cjam
I count 10 bytes
while in Jelly AẠȦẠ12345 would perfectly be 9 bytes
@EriktheOutgolfer TIO doesn't actually check that the input can indeed be represented in the SBCS of the currently active language. It is up to the user to not "cheat".
not what I meant, but cjam should have utf8 encoding set instead of sbcs since cjam just doesn't work like that
@EriktheOutgolfer Ah, that's probably a mistake. You can submit a pull request.
 
3 hours later…
13:47
@feersum ELF, e.g., takes an xxd-style hexdump as code.
@EriktheOutgolfer That's false. CJam supports all encodings that are installed on your OS.
@Dennis oh...so you have sbcs for chars 128-255 or something? hmm...
TIO's byte counter is a bit primitive right now. X bytes (SBCS) means that it's X bytes long if it can be encoded in a SBCS. Note that it doesn't even specify which one.
yeah that's what I meant
But CJam supports all of them. Just set LANG to the appropriate value and CJam will use that encoding.
14:16
@Dennis We just released Rain-Flak 1.5.1, it patches a few bugs in 1.5.0. Could you pull to TIO?
@WheatWizard oh rain in the brain
@Dennis Any progress on QuadR and QuadS?
@Adám I was waiting for you to commit quadr and quads. Apparently, that happened 14 hours ago. >_>
14:36
or even shouldiblamelaziness.com and start pressing the f5 key more often
GitHub auto-refreshes. Looks like I simply didn't check after 20:00.
@Dennis if you go to another page and then back to it that might not be the case though...but your case was ^^ it seems
Call me lazy if you want to, but I simply stopped working on TIO at 20:00 yesterday.
Contrary to what some people want to believe, I'm only human and I need some time off every now and then.
17
@Dennis that defies common ppcg assumptions :P
@Adám Almost good to go. It would be nice if you could mark the scripts quadr and quads executable.
14:51
@Dennis I don't know how to do that.
If your filesystem doesn't have Unix permissions, you can do git update-index --chmod=+x quadr from the command line (and the same for quads).
I can't chmod locally since the next pull would either fail or overwrite the changes.
@Dennis Thank you. Like so?
@Dennis Yay! Thank you.
@WheatWizard I can't find that language on TIO.
15:01
@Dennis Thats brain-flak
Right. Pulled.
Thanks! :)
@Dennis Could you pull Husk, please? The interpreter should now be callable from a separate folder.
@@Dennis I left a dummy string in :-( Can you pull?
@Zgarb It is. :) Thanks!
@Adám Done.
15:08
@Dennis Thanks. Somehow, I can't add Arguments. The section opens, but clicking the "+ add" doesn't do anything. Not just for QuadRS, also for all other languages.
Uh oh, when/how did that happen?
Fixed. Sorry about that.
@Dennis what was it? I did a hard refresh...maybe you fixed it before?
Double encoding, sort of. I made a low-priority edit the other day to remove some code than required ES6 features (it stilled work in ES5, but it created a global variable by accident). I replaced a more or less complicated invocation with a function I had already made for that purpose, but I ended up calling $(...) twice in the process...
@Dennis ooh creepy
15:21
\o/
 
1 hour later…
Okx
Okx
16:40
@Dennis Mind adding Neim to TIO? I've been having a problem with the online interpreter. I'll try not to bug you to update it :P (link to jar: github.com/okx-code/Neim/tree/master/build/libs run java -jar <JAR> <FLAGS> <FILE>, hello world: (Hello, World!)B )
@Okx Will build/libs always contain exactly one JAR file?
Okx
Okx
yes
16:58
@Okx That trailing ) confused me.
Okx
Okx
haha
Okx
Okx
neat
 
2 hours later…
18:38
@Dennis can you pull Chip please?
19:03
Also, I have a question about buffered output when the 60-second limit is reached. Is it possible to capture the buffer contents at that point or is that impossible in TIO?
19:22
@Dennis Could you pull QuadRS?
19:47
@Phlarx Done.
@Dennis Thanks,
@Phlarx All scripts in TIO's backend are unbuffered, but Python itself is buffering the output.
@Dennis I see. In my experience, when I use python's buffered output, and hit ^C, it'll dump out what it has. Maybe that's a shell or distro feature?
Or TIO's using a different means to terminate the process?
@Dennis (also, thanks!)
20:56
@Phlarx TIO send SIGINT, but Python probably behaves differe ntly in a terminal.
@Dennis Any way to have Code have every line end with a newline, or should I take care of that myself?
I'm not sure what you mean. Is this about the last line?
@Dennis Yes.
You'd have to account for that in your wrapper. Preferably without modifying the actual source file.
@Dennis I don't know enough bash to do that, but I could handle it on the APL side.
21:48
@Dennis If this makes sense, then feel free to pull.
Leo
Leo
@Dennis pull Husk please?
22:05
@Adám That will append a newline even if the program ends with a nrwline. Is that a problem?
@Dennis No, that should be fine if my calculations are correct.
22:25
@Leo Done.
@Dennis Thank you.
@Adám Alright, pulled.
@Dennis SE does not handle trailing newlines correctly in <pre><code> blocks. You need to insert <br>.
Looks fine to me.
22:42
@Dennis Yeah, but if you make a post with the trailing newline, SE will not render it (well, it will initially in the preview, but not if you post)
No repro. Do you see the newline in this post?
8
A: Appending Numbers

DennisRetina, 5 bytes \d*¶ Takes two strings as input, separated by a newline. Try it online!

SE didn't alter the HTML. In what browser are you having problems?
@Dennis Chrome 59/Win10. Do you see the newline in this post?
0
A: Get rid of whitespace on otherwise empty lines

AdámQuadR, 5 bytes \s+$ Try it online! \s+ replace one or more whitespace chars followed by $ end of line  with nothing

No. It's the syntax highlighting.
@Dennis The syntax highlighting removes newlines‽
I saw it for a second, and then it went away
22:47
Yes, that's when syntax highlighting kicked in.
I thought that only changed colors.
Changing colors requires wrapping tags around things to color them. Part of that seems to have gone wrong.
I guess I should file a bug report then.
The questions is where. SE uses a third-party tool for this.
Google to blame.
22:55
It's different on Mother Meta.
0
A: Formatting Sandbox

Dennistrailing newline 1 trailing newline 2 trailing newline 3 trailing newline 4

If I use newline plus <br> there, I get two newlines. Without <br> no newline gets rendered at all.
23:45
@Dennis Is there a way to link to the TIO descriptive page, the one I see when clicking TIO in the top left corner?

« first day (309 days earlier)      last day (2185 days later) »